污水处理可以控制水污染，同时提供再生水作为非常规水资源。硫自养反硝化技术因运行成本低、污泥产量低等优势受到关注，已逐步用于污水处理厂深度脱氮。本研究针对硫自养反硝化启动时间长、脱氮速率低，通过调控硫自养反硝化过程中多硫化物的生成，缩短反应器启动时间、提高反应器脱氮速率，为硫自养反硝化技术的实用化提供理论支持。 研究牛血清蛋白对脱氮硫杆菌反硝化性能的影响，发现牛血清蛋白加入硫粉中和加入生物膜中脱氮速率分别为18.7 mg-N/(L·d)和8.3 mg-N/(L·d)，牛血清蛋白加入硫粉中具有更优的脱氮速率。1%质量比牛血清蛋白掺混硫片脱氮速率为18.7 mg-N/(L·d)，是对照组的2.6倍，反硝化4 d后多硫化物浓度为11.9 mg/L，是对照组的3.1倍，牛血清蛋白促进单质硫转化为多硫化物，提高了脱氮速率。连续实验发现，1%质量比牛血清蛋白掺混的单质硫填充床启动时间为13 d，比对照组缩短2 d，启动后脱氮速率为38.3 mg-N/(L·d)，是对照组的1.5倍。 利用硫化物与单质硫反应生成多硫化物，提高反应器脱氮速率。反应器中加入硫化钠（80 mg/L）后脱氮速率为9.8 mg-N/(L·d)，是对照组的3.2倍。反应器中加入不同浓度的硫化钠，随着硫化钠浓度由40 mg/L增加到160 mg/L，反应1 d后生成的多硫化物浓度由16.7 mg/L增加到36.9 mg/L，反应器脱氮速率由7.3 mg-N/(L·d)增加到11.1 mg-N/(L·d)。硫化钠与单质硫生成多硫化物，提高了硫氧化菌对单质硫的利用，提高了脱氮速率。 在含有硫粉的SBR反应器中培养厌氧污泥生成含有蛋白质的生物膜，将其加入硫填充床反应器中，10 d后反应器接种完成，启动时间比对照组缩短了10 d，启动成功后水力停留时间为24 h时反应器脱氮速率为101.4 mg-N/(L·d)，是对照组的2.1倍；采用上述接种方法的反应器在15℃低温下，水力停留时间为3 h的脱氮速率为134.4 mg-N/(L·d)，是对照组的1.8倍。在硫填充床反应器启动时加入160 mg/L的硫化钠，12 d后接种完成，比对照组缩短6 d，启动成功后水力停留时间为2 h时反应器脱氮速率为346.9 mg-N/(L·d)，是对照组的1.4倍；反应器在15℃低温下，水力停留时间为2 h的脱氮速率为107.4 mg-N/(L·d)，是对照组的1.5倍。

Wastewater treatment is an important technology to control water pollution and provide reclaimed water as additional water resource. Sulfur autotrophic denitrification has received increasing attention due to its low operating cost and low sludge output, and it has been applied in practical nitrogen removal in wastewater treatment plants. In order to address the long start-up time and low nitrogen removal rate in sulfur autotrophic denitrification, polysulfides was prepared and employed with elemental sulfur in sulfur autotrophic denitrification. Experimental results revealed that the reactor’s start-up time was shortened and nitrogen removal rate was increased. The effect of bovine serum albumin (BSA) on denitrification performance was studied. With additional BSA in sulfur powder and Thiobacillus biofilm, the nitrogen removal rate was 18.7 mg-N/(L·d) and 8.3 mg-N/(L·d), respectively. Different mass ratios of BSA in sulfur including 1%, 2% and 4% were tested, the corresponding nitrogen removal rate was 18.7 mg-N/(L·d), 24.0 mg-N/(L·d) and 26.8 mg-N/(L·d), respectively, 2.6, 3.3 and 3.7 times higher than that of the control group (without additional BSA). After 4 days of denitrification with sulfur modified by BSA, the polysulfide concentration was 11.9 mg/L, 3.1 times of the control group. Long-term experiment of two weeks found that the start-up time of elemental sulfur packed bed with the presence of BSA was 13 days, 2 days shorter than the control group. The denitrification rate after start-up reached 38.3 mg-N/(L·d), 1.5 times higher than that of the control group. Polysulfides can be formed by the reaction of sulfides with elemental sulfur, through which sulfur autotrophic denitrification could be accelerated. Results showed that adding sodium sulfide (80 mg/L) increased the denitrification rate to 9.8 mg-N/(L·d), 3.2 times that of the control group (without additional sodium sulfide). Different sodium sulfide concentrations were added to the reactor, and with the increase of sulfide concentration from 40 mg/L to 160 mg/L, the concentration of polysulfide generated after 1 day of reaction increased from 16.7 mg/L to 36.9 mg/L, and the denitrification rate of the reactor increased from 7.3 mg-N/(L·d) to 11.1 mg-N/(L·d). Sodium sulfide reacts with elemental sulfur to form polysulfide, which improves the utilization of elemental sulfur by sulfur oxidizing bacteria and improves the nitrogen removal rate. This study also tested a new method of inoculating the elemental sulfur packed bed, where a protein-containing biofilm was prepared and employed as an inoculum source. The reactor was started up in 10 days using this method, while the start-up time in control group (normal start-up process) was 20 days. The denitrification rate of the experimental reactor was 101.4 mg-N/(L·d), 2.1 times of the control. The denitrification rate of the reactor with the above inoculation method at a low temperature of 15°C was 134.4 mg-N/(L·d) with a hydraulic residence time of 3 h, which was 1.8 times that of the control group. Further adding sodium sulfide (160 mg/L) into the experimental reactor increased the denitrification rate to 346.9 mg-N/(L·d), which is 1.4 times that of the control. The denitrification rate of the experimental reactor at a low temperature of 15℃ and a hydraulic retention time of 2 hours after successful inoculation was 107.4 mg-N/(L·d), which is 1.5 times that of the control.